Self-supervised learning for classifying paranasal anomalies in the maxillary sinus - Takeaways - MDSpire

Self-supervised learning for classifying paranasal anomalies in the maxillary sinus

  • By

  • Debayan Bhattacharya

  • Finn Behrendt

  • Benjamin Tobias Becker

  • Lennart Maack

  • Dirk Beyersdorff

  • Elina Petersen

  • Marvin Petersen

  • Bastian Cheng

  • Dennis Eggert

  • Christian Betz

  • Anna Sophie Hoffmann

  • Alexander Schlaefer

  • June 8, 2024

  • 0 min

Share

  • 1

    Self-supervised learning (SSL) is proposed to classify paranasal anomalies in the maxillary sinus using unlabelled MRI data.

  • 2

    A 3D convolutional autoencoder (CAE) is utilized to reconstruct maxillary sinus volumes and localize anomalies through reconstruction errors.

  • 3

    The study emphasizes the importance of accurately diagnosing paranasal anomalies to prevent patient distress and reduce healthcare costs.

  • 4

    The method leverages labelled healthy maxillary sinus data to enhance downstream classification performance of anomalies.

  • 5

    Post-processing strategies and loss functions are investigated to improve feature discrimination in the self-supervised learning task.

Original Source(s)

Self-supervised learning for classifying paranasal anomalies in the maxillary sinus

  • by Debayan Bhattacharya, Finn Behrendt, Benjamin Tobias Becker, Lennart Maack, Dirk Beyersdorff, Elina Petersen, Marvin Petersen, Bastian Cheng, Dennis Eggert, Christian Betz, Anna Sophie Hoffmann, Alexander Schlaefer

  • June 8, 2024

Related Content

CRS Trial Reveals Surgery Divide

MACRO qualitative study explores why patients differ in attitudes toward endoscopic sinus surgery.

Who Decides When a Doctor Steps Back?

Cognitive decline, credentialing, and why “I’ll know” often isn’t enough.

Medical Oddities: For the Birds—and the Brain

Birds for brains, blackout laughs, anti-aging tea, thin air, muscle memory—and torn marriages, all with a clinical touch this week.